KI fuer mehr Arbeitssicherheit: Praevention im Bauwesen durch digitale Innovation

Summary

In-depth interview with Ellen Hellmann, BG BAU project leader for Digital Transformation, about the AI-based accident prevention system. Provides detailed technical information: Random Forest model selected from 60 tested models, trained on 10 million data points using open-source technologies, risk scores 0-1 with traffic-light display, weekly updates. Describes bias mitigation (normalisation by employee count), voluntary usage principle, project team (BG BAU, BG-Phoenics, Accenture), and follow-up project BeA. Primary source with most technical detail.